A Comparison of Classifiers for Solar Energetic Events

Abstract

We compare the results of using a Random Forest Classifier with the results of using Nonparametric Discriminant Analysis to classify whether a filament channel (in the case of a filament eruption) or an active region (in the case of a flare) is about to produce an event. A large number of descriptors are considered in each case, but it is found that only a small number are needed in order to get most of the improvement in performance over always predicting the majority class. There is little difference in performance between the two classifiers, and neither results in substantial improvements over simply predicting the majority class.

The Flare Likelihood and Region Eruption Forecasting (FLARECAST) Project: Flare forecasting in the big data & machine learning era

“…We note at this point that the relatively small number of best predictors found by Florios et al (2018) and Campi et al (2019) essentially aligns with earlier results by Ahmed et al (2013); Bobra and Couvidat (2015) and Barnes et al (2017). The latter study also found that a set of three or four predictors was sufficient to guarantee best performance for two classifiers, NPDA and RF.…”

Forecasting Solar Flares Using Magnetogram-based Predictors and Machine Learning

“…This is one of the first times RF is used for flare forecasting. Other related works are Liu et al (2017) and Barnes et al (2016). Furthermore, three recent applications of RF in astrophysics are by (Vilalta, Gupta, and Macri, 2013;Schuh, Angryk, and Martens, 2015;Granett, 2017).…”

“…More often, the techniques used by researchers were: neural networks (Wang et al, 2008;Yu et al, 2009;Colak and Qahwaji, 2009;Ahmed et al, 2013), support vector machines (Li et al, 2008;Yuan et al, 2010;Bobra and Couvidat, 2015;Boucheron, Al-Ghraibah, and McAteer, 2015), ordinal logistic regression (Song et al, 2009), decision trees (Yu et al, 2009) and relevance vector machines (Al-Ghraibah, Boucheron, and McAteer, 2015). Very recently, random forests have also been used (Barnes et al, 2016;Liu et al, 2017).…”